Hybrid bat inspired algorithm for multiprocessor real-time scheduling preparation
N. B. Arunekumar, A. Selva Kumar, K. Suresh Joseph · 2016
Real Time Processor Scheduling with no preemption of tasks is a class NP-hard problem. We have attempted to get the best task allocation schedule for sporadic tasks, such that all the tasks are being scheduled without missing its deadline. Bat algorithm was proposed from the motivation behind the bat's echolocation behaviour. Bat Searches for the prey in the given search location once found it tries to catch it by converging towards the direction of the prey. The existing system comprises of multiprocessor scheduling using bat algorithm which has been proved to be very efficient. Based on it we have proposed a modernistic algorithm to solve the Multi-objective multiprocessor scheduling algorithm of a soft real time scheduling system for both periodic and sporadic tasks that is completely scalable in nature. We have considered two objectives which are the maximum processor utilisation and minimised deadline miss. We have conducted experiments with several instances and found out that we are able to schedule the tasks by balancing both the objectives with the probability of 100% in arriving to the optimised schedule of the given tasks.